Single cell transcriptomes reveal characteristics of miRNA in gene expression noise reduction

Isogenic cells growing in identical environments show cell-to-cell variations because of stochastic gene expression. The high level of variation or noise could disrupt robust gene expression and result in tremendous consequences on cell behaviors. In this work, we showed evidence that microRNAs (miRNAs) could reduce gene expression noise in mRNA level of mouse cells based on single-cell RNA-sequencing data analysis. We identified that miRNA expression level, number of targets, targets pool abundance and interaction strength of miRNA with its targets are the key features contributing to noise repression. MiRNAs tend to work together as cooperative sub-networks to repress target noise synergistically in a cell type specific manner. Using a physical model of post-transcriptional regulation, we demonstrated that the accelerated degradation with elevated transcriptional activation of miRNA target provides resistance to extrinsic fluctuations. Together, through the integration analysis of single-cell RNA and miRNA expression profiles. We demonstrated that miRNAs are important post-transcriptional regulators for reducing gene expression noise and conferring robustness to biological processes.